An evaluation of the practice of transparency and reproducibility in addiction medicine literature.

ABSTRACT

BACKGROUND: Credible research emphasizes transparency, openness, and reproducibility. These characteristics are fundamental to promoting and maintaining research integrity. The aim of this study was to evaluate the current state of reproducibility in the field of addiction science. DESIGN: Cross-sectional design. MEASUREMENTS The National Library of Medicine catalog was searched for all journals using the subject terms tag Substance-Related Disorders [ST]. Journals were then searched via PubMed to identify publications from January 1, 2014, to December 31, 2018; 300 publications were randomly selected from among those identified. A pilot-tested Google form containing reproducibility/transparency characteristics was used for data extraction in a duplicated and blinded fashion by two investigators. FINDINGS: Slightly more than half of the publications were open access (152/300, 50.70%). Few publications had pre-registration (7/244, 2.87%), material availability (2/237, 0.84%), protocol availability (3/244, 1.23%), data availability (28/244, 11.48%), or analysis script availability (2/244, 0.82%). Most publications provided a conflict of interest statement (221/293, 75.43%) and funding sources (268/293, 91.47%). One replication study was reported (1/244, 0.4%). CONCLUSION: Our study found that current practices that promote transparency and reproducibility are lacking, thus, there is room for improvement. In particular, investigators should pre-register studies prior to commencement. Researchers should also make the materials, data, and analysis script publicly available. Further, individuals should be transparent about funding sources for the project and financial conflicts of interest. Research stakeholders should work together toward improvements on these matters. With such protections, the field of addiction medicine can better disseminate the information necessary to treat patients.